The long term objective of our research is to improve the treatment of bone cancer through development of animal models resembling human disease and through testing new experimental therapies in these models. This competitive renewal application builds on exciting progress made during the previous funding period. That progress demonstrated several critical points: 1) osteoclast lineage cells can be used as an effective gene delivery system to sites of bone cancer; 2) osteoclast precursor cells serve as the ideal gene delivery system, in that the Good Samaritan effect is not present, permitting the gene delivery vehicle (osteoclast precursor) to survive the therapy and therefore provide extended expression of the therapeutic molecule cytosine deaminase. Our Preliminary Studies now show that lentiviral approach for gene delivery is operational in our hands and is very promising as a vector for providing the therapeutic gene to sites of tumor. The current proposal adopts modifications of the osteoclast-directed treatment scheme that was studied during the previous funding period. These adaptations will advance our work toward a clinical trial in patients. The proposed modifications will increase the specificity of the therapy, will permit regulation of gene expression, and will decrease the toxicity of current treatment. These modifications include use of a new therapeutic vector (lentivirus base) and adaptation of the osteoclast delivery to the method of direct intratumoral injection. The feasibility of the approach and the experimental methods and design are greatly enhanced in this A2 submission, as all concerns raised by the reviewers have been addressed. Specifically, we have demonstrated effective therapeutic gene transduction into osteoclast precursor cells using a lentiviral approach. We have demonstrated that hypoxia induces HRE activation in purified populations of osteoclast lineage cells in vitro, and we have demonstrated that bisphosphonate therapy does not interfere with the survival of intratumoral-injected osteoclast progenitor cells. The aims of this application mirror those previously proposed, but are more strongly supported by our response to the review. Aim 1 will determine if hypoxia and radiation- activated promoters can mediate osteoclast-directed killing of bone cancer cells in vitro. In this aim, osteoclast progenitor cells will be transduced with lentivirus containing our molecular switch vector which is responsive to hypoxia and radiation, and which drives expression of the enzyme cytosine deaminase. The cytosine deaminase prodrug enzyme system will then be tested in vitro. Aim 2 will determine if Tg osteoclast precursor cells expressing the prodrug enzyme cytosine deaminase can kill bone cancer in vivo, and will determine if authentic osteoclast precursor cells transduced with the lentivirus molecular switch cytosine deaminase vector can kill bone cancer in vivo. Successful completion of this aim will serve as a next step toward experimentation in humans. Aim 3 will determine if the 5' portion of the TRAP gene regulates expression of the TRAP 1C transcript. This aim is critically important, as it will help direct future design of osteoclast-specific gene delivery vectors, ensuring cell-type-specific expression and maximizing safety.